Who pays for lost inventory at a 3PL?

Then the warehouse says: “Within the shrinkage allowance.”

That sentence can erase margin fast.

The short answer: the brand usually pays for lost inventory up to the contractual shrinkage allowance. Beyond that threshold, the 3PL may pay—but usually only if the contract assigns liability or you can show negligence. And even then, the reimbursement is commonly based on your wholesale cost or COGS, not the retail revenue you expected to collect.

This is the part operators routinely underestimate. A 3PL inventory shrinkage allowance standard is not a minor legal footnote. It is a direct variable in gross margin, inventory planning, cash conversion, and your ability to scale acquisition without creating a fulfillment leak behind the scenes.

I have seen teams fight over a few thousand “missing” units while continuing to scale campaigns that send more product into the same broken operation. Wrong order. Stop the loss first. Then scale.

The mechanics of shrinkage allowances: the loss your brand is expected to absorb

Inventory shrinkage is the gap between what the warehouse management system says should be in stock and what a physical count finds. It can include mis-picks, damaged units, receiving errors, misplaced cartons, theft, unrecorded disposals, and plain old bad scanning discipline.

Most 3PL agreements do not make the provider responsible for every unit that disappears. They include a shrinkage allowance—a predefined amount of loss the merchant absorbs before the provider owes reimbursement.

The typical range is 0.5% to 1% of on-hand inventory. That is the commercial baseline many brands walk into without doing the math.

Say you hold 100,000 units at a 3PL, with a landed COGS of $12 per unit:

Now layer in reality. That $24,000 is not just inventory cost. Those units may have been allocated to a paid acquisition push, a creator drop, a retail replenishment order, or a subscription cohort. The warehouse loss can trigger stockouts. Stockouts kill conversion rate. Lower conversion means CAC rises even when your media buyer changes nothing.

That is why “acceptable inventory shrinkage rate 3PL” is not a procurement question. It is a growth question.

Your shrink allowance is not theoretical loss. It is a pre-approved dent in contribution margin.

The definition matters as much as the percentage. I want the contract to state, in plain language:

• Whether the allowance applies to average inventory, monthly ending inventory, annual inventory, or another calculation base.
• Whether damage, theft, mis-shipments, returns processing errors, and stock adjustments all count toward the same allowance.
• When a discrepancy becomes “confirmed”—at the moment of a WMS adjustment, after cycle-count verification, or only after an annual physical count.
• Whether losses tied to documented 3PL errors sit outside the allowance.
• Whether the allowance resets annually, quarterly, by facility, or by SKU category.

A vague 1% allowance is an invitation to a margin dispute. A clear 1% allowance is still a cost, but at least it is a forecastable cost.

Liability starts after the threshold—but negligence changes the conversation

The usual 3PL lost inventory policy works in layers.

First, the merchant absorbs losses within the contractual shrinkage allowance. Second, losses beyond that cap may trigger reimbursement. Third, the 3PL’s liability provisions can limit what it actually pays. This is where a lot of founders discover that “liable” and “making us whole” are not remotely the same thing.

In the United States, warehouse operators must exercise “reasonable care” over stored goods under the Uniform Commercial Code. If a provider’s negligence causes the loss, that standard can support a claim against the warehouse.

But do not treat “reasonable care” as a magic phrase that instantly wins an argument. It is fact-specific. You need records: receiving scans, bin movements, cycle-count history, order exceptions, damaged-goods logs, camera footage where available, and a timeline showing where the process failed.

The real fight is rarely over whether an item is missing. The fight is over why it went missing.

Here is how the distinction usually plays out:

I push operators to stop asking, “Will the 3PL cover it?” Ask a sharper question: “What evidence do we need to prove the loss sits outside the allowance and inside their operational failure?”

That question forces discipline before the problem becomes expensive.

A provider with a clean WMS, barcode-enforced receiving, location-level scan history, documented cycle counts, and rapid discrepancy workflows can answer it. A provider running spreadsheet patches, delayed adjustments, and “we’ll investigate” emails cannot.

If your warehouse cannot trace a unit from inbound receipt to storage location to pick confirmation to outbound manifest, you do not have supply-chain visibility. You have hope with a dashboard.

Reimbursement reality: they usually pay COGS, not your retail price

Here is the moment that stings: even when the 3PL accepts responsibility, reimbursement commonly lands at wholesale cost or COGS, not retail value.

That means a product that retails for $68 may be reimbursed at a $14 landed product cost. The missing $54 is not “extra.” It is the revenue, gross-profit potential, paid-media recovery, and future LTV you expected that unit to create. The 3PL generally is not underwriting your growth model.

This is especially brutal for brands with high AOV, premium packaging, or aggressive CAC targets. Beauty, supplements, specialty food, apparel capsules—these categories can carry a large gap between production cost and retail value. A warehouse liability cap turns that gap into your risk.

Consider a beauty brand relaunching its retention engine. The marketing side may be rebuilding demand with better product storytelling, community, and recurring purchase behavior—the same broad pressure behind Glossier’s plan to reignite its marketing magic. But fulfillment has to hold up its end. Losing a hero SKU at the warehouse does not merely cost one unit’s COGS; it can break a replenishment moment and hand a repeat customer to a competitor.

Many 3PL agreements also limit liability by unit, pound, package, or shipment. Those clauses can produce a reimbursement figure that bears little resemblance to your actual loss, particularly for small, high-value goods.

Read the cap before you sign. Then run your worst SKU through it.

If the contract says the 3PL will pay a fixed amount per pound, calculate what that means for a lightweight serum, a jewelry item, a premium electronic accessory, or a limited-edition bundle. The answer can be ugly.

A liability clause that looks harmless at pallet scale can become absurd at SKU scale.

I do not accept vague language such as “3PL shall reimburse inventory value.” Define the value. Is it purchase cost? Landed COGS? Replacement cost? Is freight included? Are duties included? Does custom packaging count? What documentation must the merchant produce? How fast must the 3PL issue a credit?

Every undefined term becomes a delay. Delays become stockouts. Stockouts become a performance-marketing problem by Friday afternoon.

Benchmark the operation, not the sales pitch

A 3PL can call its shrinkage “normal” all day. Your job is to benchmark the claim against its controls and your economics.

Basic warehouse operations often operate in a 1% to 3% shrinkage range. Across U.S. 3PL facilities, reported average shrinkage has been around 1.4%. That is not a comfort number. It is a warning that average performance can still be expensive.

A modern target for inventory accuracy is 99.5% or better. If your provider is consistently over 2% shrinkage, I would not treat that as background noise. I would treat it as a control failure requiring a commercial reset: deeper audit rights, tighter service-level terms, a revised allowance, or a provider change.

At the other end, highly automated fulfillment operations can report shrinkage below 0.01%. That does not mean every merchant needs robotics tomorrow. It means the old excuse—“some loss is unavoidable”—needs context. Some loss is normal. Uncontrolled loss is not.

The difference is process density:

1. Inbound receives are scanned at the right granularity. Case-level receiving might work for uniform commodity inventory. It fails when cartons contain mixed SKUs, bundles, lots, or high-value units. The more expensive and varied the catalog, the more SKU-level precision you need.

2. Every inventory move creates a system event. Putaway, replenishment, bin transfer, pick, pack, damage quarantine, return disposition—each move should leave a timestamped trail. If inventory can move without a scan, it can disappear without a defensible explanation.

3. Cycle counts target risk, not convenience. Your A movers, high-value SKUs, return-heavy items, and bundle components deserve more frequent checks than slow-moving basics. Quarterly counts are common, but the cadence should follow exposure.

4. Annual physical reconciliation closes the loop. Quarterly cycle counts catch drift. A full annual physical inventory reconciliation establishes whether the operation’s reported stock can survive a hard count. You need both.

5. Root-cause codes drive action. “Adjustment” is not a root cause. “Inbound overage unverified,” “pick-face mis-slot,” “unscanned damage,” and “return received without SKU confirmation” are root causes. One leads to a shrug. The other leads to a fix.

I would also request performance by SKU class, not just one blended inventory-accuracy number. A 99.7% headline can conceal a disaster in the 20% of products generating 80% of revenue.

The contract is only half the defense: build an operating system around it

Brands lose leverage when they discover discrepancies months later. The 3PL has changed shifts, moved inventory, overwritten logs, or completed a physical count that buries the original failure. You need an active reconciliation rhythm—not a once-a-year panic attack.

My baseline playbook looks like this:

• Set a dispute clock. Define how many days the 3PL has to investigate and resolve a discrepancy after you flag it. Open-ended investigations protect nobody except the provider’s cash flow.
• Reconcile WMS data against your commerce and ERP records. Your inventory picture should account for available units, allocated units, damaged units, returns, inbound stock, and adjustment activity. Do not compare only “available” totals.
• Track discrepancy dollars, not just units. Fifty missing low-cost samples and fifty missing premium devices are not equivalent. Report loss by COGS, retail exposure, SKU velocity, and campaign dependency.
• Demand regular cycle-count reporting. Quarterly cycle counts combined with annual physical reconciliation are standard methods. For high-velocity or high-value inventory, run targeted counts more frequently.
• Create an exception dashboard. Watch negative inventory adjustments, repeated bin changes, damaged-unit growth, inbound variances, return hold aging, and orders short-shipped because stock was not actually where the WMS claimed.
• Insure the gap. Add a rider to your own insurance policy that lists the 3PL facility as a storage location. Do not assume the warehouse’s standard policy will cover your goods, especially in a major-loss event.

That insurance point gets skipped because it is not exciting. Neither is discovering after a flood, fire, or catastrophic loss that your operational agreement contains exclusions and the warehouse’s coverage was never designed to protect your full inventory value.

A good 3PL should not resist reasonable reporting, cycle counts, and defined claims mechanics. If they do, they are telling you exactly how painful the first major discrepancy will be.

Do not let inventory loss become invisible CAC

Lost inventory creates a sneaky reporting problem. Finance books a write-off. Operations marks an adjustment. Growth keeps optimizing ads against blended revenue data. Nobody connects the loss to acquisition efficiency.

I do.

If a hero SKU disappears, you can lose revenue from active campaigns, delay replenishment, force substitutions that reduce conversion, and damage repeat-purchase behavior. The immediate inventory cost is only the visible layer. The downstream cost can hit CTR, CVR, refund rate, subscription retention, and LTV.

That is why I want fulfillment data in the weekly growth meeting. Not every operational metric—do not drown the room—but the ones that can sabotage demand:

• Inventory accuracy on revenue-driving SKUs
• Units in damage or quarantine status
• Stockout risk for campaign-linked products
• Returns awaiting disposition
• Reconciliation discrepancies by dollar value
• 3PL claim aging and unresolved adjustments

Your paid media does not operate in a vacuum. You cannot dominate an auction while your warehouse quietly bleeds the units your ads are trying to sell.

The answer to “who pays for lost inventory at a 3PL?” is therefore not simply “the 3PL” or “the brand.” It is: the brand pays first within the negotiated allowance; the 3PL may pay beyond it or when negligence is proven; and the contract determines whether that payment resembles a real recovery or a token credit.

Read the shrinkage clause. Calculate it against your top SKUs. Stress-test the liability cap. Reconcile inventory before the quarterly count forces the conversation.

Then execute today: pull your 3PL agreement, find the shrinkage allowance and per-unit liability limit, and model the real dollar exposure before you spend another dollar scaling demand.